Site-Specific Epicardium-to-Endocardium Dissociation of Electrical Activation in a Swine Model of Atrial Fibrillation

Abstract Objectives This study sought to define the extent and spatial distribution of endocardial-epicardial dissociation (EED) in a swine model. Background The mechanisms underlying persistent atrial fibrillation (AF) remain unclear. Methods Sixteen swine underwent simultaneous endocardial and epicardial mapping using 32-electrode grid catheters. This included 6 swine with rapid atrial pacing–induced atrial remodeling. Three right atrial (RA) and 3 left atrial (LA) regions were mapped during sinus rhythm, atrial pacing, acute or persistent AF, and AF in the presence of pericardial acetylcholine. Unipolar electrogram recordings over 10-s epochs underwent offline phase analysis using customized software. Regional activation patterns on paired surfaces and the instantaneous phase at each matched electrode location were analyzed. EED was defined as paired electrodes out of phase by ≥20 ms. Results The mean distance between matched endocardial-epicardial electrode pairs was 4.4 ± 1.8 mm. During episodes of AF, rotational activations with ≥3 full rotations were not seen. EED was seen during 34.4 ± 16.4% of mapped time periods: LA > RA, persistent > acute AF in the LA, and acetylcholine-induced > acute AF in both atria (p  Conclusions Marked EED was seen in a swine model of AF, particularly during persistent AF. There was significantly more EED in the LA than the RA and, particularly, in the LA PW and LAA. Mapping approaches limited to the endocardium may not sufficiently characterize the complexity of AF.